"Increase" gate threshold voltage for HV high dv/t?

[rentner]

Hey folks, I want to build a H Bridge with high current capabilities, high speed (>200khz) short turn on/off times and a relative huge voltage for that stuff of >300V rail/-rail.

If the low side N-Fet pulls down, the high side Mosfet always turns back on and creates shoot through. This of course only happens because it is an inductive load. (Filter)



The Turn-Off diode basically does nothing for the 1V threshold - it barely conducts any current.

I think of using HF Mosfets with 6nC total gate charge for very dast and easy switching, but the tiny gate threshold voltage of 1V at this high dv/t conditions has almost no chance to stay off. Gate driver uses BJT output stage - as my fetish implies - it is discrete. The high side driver works at "light speed" and runs stable, but the Output stage has no chance to hold the gate off. I don't realy feel like Mosfet totem pole configuration because of... well shout through. The driver circuit would be way to complicated and delay switching events a ton.


How the hell do they get these Class D Amps get to work? I simulated it with real Driver ICs and at certain Q(G) and V(GS)

every high side driver IC renders useless. 300V in 50ns is the goal... Either you have an Idea, how to convert a Square PushPull Totem Pole BJT signal in to mosfets with pull down to zero volts or whatever else is possible. If you have an Idea, I am totally waiting to here it before I get totally insane!

Edit: This is the link to the LT-Spice file so you can take a look at the problem. Just simulate a few tens of microseconds and you will see the shoot through. I don't manage to fix the solution and don't 100% get the problem...

http://s000.tinyupload.com/index.php?file_id=99962263409283120830

[rentner]

Push!

Thread got to oblivion and no click in a week!

[fourtytwo42]

First it would be preferable to provide a schematic as a pdf so people can see what you are doing without resorting to downloading a file into LTspice

Second with no knowledge of your circuit I would suspect the drive impedance of the high side mosfet is to high.

[capt bullshot]

A negative voltage applied to the gate in off-state may help. This is standard with IGBTs, no so common with MOSFETs.
Or have a look of some of these optoisolated IGBT drivers from avago, some of them have special miller clamp for this issue.
If you use discrete BJTs, look for collector currents in the range of several Amps. Zetex / On make such transistors especially for gate driver purpose. Smaller transistors won't work properly here.

[Circlotron]

A negative voltage applied to the gate in off-state may help.

That's what I was going to say too.